[0001] This invention relates to a pulse width modulation (P.W.M.) system, and particularly
to such a system when employed to apply amplitude modulation to the R.F. stage of
a radio transmitter.
[0002] Such systems employ a pulse width amplifier to amplify the incoming radio signals
so that their maximum peak to peak voltage is 2V where V is the audio voltage applied
to R.F. stage when audio modulation is not present. The amplified audio signal is
therefore able to modulate the supply voltage applied to the R.F. stage between 0
and 2V volts.
[0003] A conventional pulse width amplifier when used for the above purpose would convert
the incoming audio signal to P.W.M. form in such a way that the length of each pulse
is equal to the pulse repetition period at maximum amplitude values of the incoming
signal and so that the length of each pulse is zero at minimum amplitude values. This
leads to distortion when the pulses are reconverted after amplification to amplitude
modulated form.
[0004] The obvious way of overcoming this problem is to attenuate the incoming audio signal
so that its amplitude is never sufficient to produce pulses of zero duration or of
a duration equal to the pulse repetition period. However, this leads to less than
100% modulation of the carrier. A further reason why 100% modulation might not be
obtained is that the switches 13 and 14 do not have zero resistance when. switched
on.
[0005] Efforts have been made in the past to solve this problem but have involved the use
of auto transformers or tapped chokes or other inductive components having consistent
characteristics over a required audio frequency band. Such components, have to handle
a very large power and the need for their inclusion can introduce considerable problems
of design and reliability.
[0006] This invention provides a circuit for supplying an amplitude modulated current to
a load, the circuit comprising: a first source of voltage V arranged to be applied
to the load; a pulse width modulation amplifier adapted to amplify a modulation signal
and to apply the amplified signal to the load via a D.C. blocking capacitor; and a
source of voltage, which differs from the voltage V, for supplying power to the amplifier.
[0007] By using a source of voltage which exceeds (by 6V) twice the voltage V it is possible
to obtain 100% modulation without using pulse widths approaching zero or approaching
the pulse repetition period. Under unmodulated conditions the D.C. levels at either
side of the blocking capacitor are different and the capacitor prevents circulation
currents arising from this potential difference. When modulation is present the capacitor
enables the voltage on the load to swing symmetrically about V, rather than V+dV/2.
[0008] The voltage sources are preferably separate! i.e. they are designed so that variations
in current drawn from either one do not substantially affect the other. This however
is not essential to the invention and it would be possible for one, lower, voltage
source to be provided by a suitable tapping on the other, higher, voltage source.
[0009] One way in which the invention may be performed will now be described by way of example
with reference to the accompanying schematic drawings in which:-
Figure 1 illustrates a circuit constructed in accordance with the invention and
Figure 2 shows various waveforms which appear at different parts of the circuit shown
in Figure 1.
[0010] Referring firstly to Figure 1, a first voltage source 1 has one output line connected
to a grounded conductor 2 and another output line 3 to which it applies a rectified
voltage V. The current through the line 3 is monitored by a current sensing device
4 which provides a control signal when the current exceeds a safe value. Additional
conventional sensing devices may be fitted to other parts of the circuit to provide
a like control signal in the event of other types of fault. The control signal is
operative to open contacts of a circuit breaker 5 and, in the particular systems illustrated,
to close contacts 5A. The components 4, 5 and 5A are collectively termed a "protection
device". The voltage source 1 is responsible for the production of a steady current
which passes through an audio-blocking inductor 6 and a load 7 and provides power
from which an R.F. carrier signal is derived.
[0011] The load 7 is the R.F. valve of a radio transmitter the anode of which, for correct
operation under unmodulated carrier conditions, needs to be maintained at a voltage
of V volts, this requirement being met by the power source 1.
[0012] The transmitter is amplitude modulated by superimposing an audio voltage on the anode
of the R.F. valve i.e. at point F on Figure 1. The most efficient and effective degree
of modulation is obtained when the maximum (i.e. loudest) audio signal to be handled
by the transmitter has a peak-to-peak amplitude of 2V at the R.F. valve anode, i.e.
the potential of point F varies, in accordance with the audio signal, between zero
volts (ground potential) and peaks of 2V. This is termed 100% amplitude modulation.
It is normally important in broadcasting transmitters to avoid or minimise audio distortion.
[0013] The anode voltage is modulated by an audio. signal which is introduced at point A
shown on Figure 1, this signal being represented diagrammatically at A on Figure 2.
The signal A is amplified by a pulse width modulation amplifier 8 which is supplied
with power from a second voltage source 9 which is notably not connected to the first
voltage source by a direct current discharge path. This voltage source 9 has, in the
illustrated circuit, one output line connected to the grounded conductor 2 and maintains
a rectified voltage of 2V+6V on its other output line 10. Positioned in line 10 is
a current sensor 11 which, when it senses a current in excess of that which, can safely
be delivered by the source 9 or which can safely be passed by switches 13 and 14,
to be described later, produces a control signal which operates a fast acting circuit
breaker 12 and an even faster acting switch 12A. A like control signal may be produced
by one or more additional conventional fault sensing devices elsewhere in the circuit.
The components 11, 12 and 12A in combination constitute a second protection device.
However, unlike the protection device 4, 5 and 5A which can be relatively slow to
respond to excessive currents, because the inductor 6 limits the rate of rise of current;
the protection device 11, 12 and 12A needs to be able to respond very quickly. It
is notable however that this second protection device does not handle all the power
delivered to the load since most of this is supplied from power source 1. Because
it does not handle all the power, the device 11, 12 and 12A can be designed to operate
reliably with much less difficulty than if it had needed to handle the whole power
consumed by the load.
[0014] The pulse width modulation amplifier 8 will now be described. The input signal A
is shown in Figure 2 and is composed of parts where its voltage exceeds zero and parts
where its voltage is below zero. This waveform A is converted by the pulse width encoder
P.W.E, whose construction may be of a known type, into a series of pulses B (Figures
1 and 2) the width of each pulse depending on the instantaneous voltage level of A.
Thus the pulse length is a minimum at times coincident with troughs in the waveform
A; and the pulse width is at a maximum at times when the waveform A is at a positive
peak. When the waveform A is at zero the pulse length at B is approximately half the
duration of the spaces between pulses. In this particular pulse width encoder the
maximum pulse width is a little less than the pulse repetition period to avoid distortion
which would otherwise occur if the pulses are not perfectly shaped. Similarly, the
minimum pulse width is greater than zero, also for the purpose of avoiding distortion.
[0015] In the illustrated system the pulse width encoder P.W.E. also produces an output
C which is the inverse of output B. The signals B and C are used to switch valves
13 and 14 in such a way that, when waveform B is at a high level, the valve 13 is
switched on; and, when C is at a high level, the valve 14 is switched on. When the
valve 13 is switched on it has very low resistance and so the point D is a little
less than 2V+6V volts as supplied by the power source 9. At other times, when the
valve 14 is switched on,the point D is a little above zero volts because it is effectively
switched to the earthed line 2. The voltage at point D therefore fluctuates between
just below 2V+6V volts and just above zero volts as shown at D on Figure 2. This follows
the waveform of B. The average voltage at D therefore follows the waveform A but is
greatly amplified.
[0016] In an alternative embodiment (not illustrated) only the signal C is generated by
the pulse width encoder, and applied to valve 14 as in Figure 1. The grid control
circuit of valve 13 is connected to the anode of valve 14 in such a way that the two
valves 13 and 14 conduct in anti-phase. There are several known methods of performing
this.
[0017] An energy store in the form of an inductor 15 is used to provide a voltage E whose
value represents the short term average voltage at D. Further energy storage components,
of conventional type, may be added to improve the degree to which the pulses are filtered
out. With no audio signal at A the D.C. voltage at E is in excess of the voltage V
by approximately 6V/2 but no D.C. current flows through inductor 15, such current
being prevented by capacitor 16. Where an audio signal of maximum modulation is applied
at A the voltage E varies between about δV/2 and 2V+dV/2 in response to variations
in the average voltage at D. Thus the variation in voltage E can equal 2V even though
100% pulse width modulation (with its associated distortion) has not been employed.
The 2V voltage variation at E is applied to the load 7 through the capacitor 16, thus
modulating the load voltage between 0 and 2V as shown at F.
[0018] In the circuit shown in Figure 1 the first voltage source 1 is required to supply
a steady current to the load 7. However, this is not essential and the invention is
applicable to systems having a controlled carrier level.
[0019] In another modification the amplifier 8 is not connected to the earthed conductor
2, the part of the line 2 shown is broken lines being omitted. Instead, the line 10
may be earthed or an earth connection may be made to a centre tap on the voltage source
9, also as shown in broken lines. An advantage of such a centre tap is that the maximum
voltage between earth and either output of the voltage source is just a little over
V volts instead of over 2V volts.
1. A circuit for supplying an amplitude modulated current to a load, the circuit comprising:
a first source of voltage V arranged to be applied to the load; a pulse width modulation
amplifier adapted to amplify a modulation signal and to apply the amplified signal
to the load via a D.C. blocking capacitor; and a second source of voltage for supplying
power to the amplifier, characterised in that the second source of voltage differs
from the voltage V.
2. A circuit according to claim 1 characterised in that the second source of voltage,
exceeds twice the voltage V.
3. A method of operating a circuit in accordance with claim 2 in which the amplitude
of modulation of said signal produces less than 100% pulse width modulation in said
amplifier but in which the voltage produced by said second source enables the amplified
signal to produce substantially 100% modulation of said voltage V applied to the load.